On Tue, 03 Oct 2006 21:30:17 +0200, Osvaldo La Rosa <[EMAIL PROTECTED]> wrote: > Hi, on a very recent PC - Intel D945GNT with ALSA as sound system- the > current saytime package talks like this: > "the time is s, nine n, twenty y pm, and d twenty six seconds" > > It seems that this should be the driver (cf. hda_intel): turning the chipset > off in the bios and replacing the audio chip with a SBLive emu10k1 results > in resolving that problem, but under ALSA saytime is slower to talk than > under OSS. Sorry, I can't help; I don't have a ALSA system.
> I'd like to see better samples in it, talking a little bit faster (more > human), and fixing each slowness problem of it. Yes, I agree the samples are not very good. Please tell me if you have better _free_ samples (their copyright must be clear). > I also not understand while I can play wav or mp3 files even under oss or > alsa, while using saytime causes so much differences between oss or alsa, + > so much quality deterioration depending on the driver who is used by alsa or > oss: > maybe I can sugges to build a unique au file who is simply sended to the > dsp, just as timidity turns a midi file into audio sended to the dsp: its > fluent, it's clear, and dealy times are becoming smaller ans smaller since > pcs are currently over 3gHz of speed. The upstream version does this; it sends the .au files to /dev/audio with the write() function. The Debian version uses sox because cat foo.au > /dev/audio doesn't work on all systems. -- Oohara Yuuma <[EMAIL PROTECTED]> Wee ki ra chs morto tes whou gauzewiga, en yehar omnis tes whou gyen. --- Raiden Freaks 2 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]